The proteins encoded by the rbs operon of Escherichia coli: I. Overproduction, purification, characterization, and functional analysis of RbsA.

The nucleotide-binding component of the high-affinity ribose transport system of Escherichia coli, RbsA, was overproduced from a T7-7 expression vector, and the protein was purified. Biochemical analyses of the purified protein indicated that the ATP analogues, 5'-FSBA and 8-azido ATP, covalently labeled the protein, a reaction that was inhibited by ATP, but not by GTP or CTP. The pure protein exhibited low-level ATPase activity with a K(m) of about 140 microM. Analyses of bacterial strains carrying chromosomal deletions of rbsA and other rbs genes suggested that RbsA is important for the chemotaxis function, a surprising result that was not anticipated from previous studies. However, an inconsistency between the several results from deletion strains raises questions regarding the interpretations of the in vivo data.

The proteins encoded by the rbs operon of Escherichia coli: II. Use of chimeric protein constructs to isolate and characterize RbsC.

Chimeric genes encoding full-length copies of rbsA and rbsC connected by segments coding for short bridge peptides were constructed and expressed in Escherichia coli. Surprisingly, the chimeric genes complemented the strain in which rbsA and rbsC were deleted. The chimeric proteins were overproduced, and the products were purified by affinity chromatography. In order to obtain highly purified protein, a poly-His leader peptide was incorporated so that Ni-chelate affinity chromatography could be employed. The leader peptide and the bridge peptide were designed with factor Xa-cleavable sites to permit recovery of the individual RbsA and RbsC protein. A rbsC gene encoding a poly-His leader was also constructed and expressed. Both the chimeric RbsA-C species and the poly-HisRbsC were produced at levels that permitted isolation of the equivalent of milligram quantities of RbsC per liter of culture. This is a substantial increase in amounts from any previous RbsC production vectors. All proteins from the rbs operon have now been overproduced and substantially purified.

Probing protein-protein interactions. The ribose-binding protein in bacterial transport and chemotaxis.

A number of mutations at Gly134 of the periplasmic ribose-binding protein of Escherichia coli were examined by a combined biochemical and structural approach. Different mutations gave rise to different patterns of effects on the chemotaxis and transport functions. The smallest residue (alanine) had the least effect on transport, whereas large hydrophobic residues had the smallest effect on chemotaxis. Comparison of the x-ray crystal structure of the G134R mutant protein (2.5-A resolution) to that of the wild type (1.6-A resolution) showed that the basic structure of the protein was unaltered. The loss of chemotaxis and transport functions in this and similar mutant proteins must therefore be caused by relatively simple surface effects, which include a change in local main chain conformation. The loss of chemotaxis and transport functions resulting from the introduction of an alanine residue at position 134 was suppressed by an additional isoleucine to threonine mutation at residue 132. An x-ray structure of the I132T/G134A double mutant protein (2.2-A resolution) showed that the changes in local structure were accompanied by a diffuse pattern of structural changes in the surrounding region, implying that the suppression derives from a combination of sources.

Identical mutations at corresponding positions in two homologous proteins with nonidentical effects.

The x-ray structure of a mutant (Gly72 to Asp) of the Escherichia coli ribose-binding protein with altered transport function has been solved and refined to 2.2-A resolution with a conventional R-factor (R-factor = [formula: see text]) of 16.0% and good stereochemistry. Comparison with the wild type ribose-binding protein shows that the structure is disturbed little at the actual mutation site, but quite appreciably in a neighboring loop. Changes in the surface of the protein at the site of mutation, however, seem to explain the functional effects. A corresponding mutation of the related glucose/galactose-binding protein has different structural and functional effects due to the different structural context of the mutation site in that protein. These results are consistent with the concept that these proteins have slightly different ways of interacting with the membrane components in transport and chemotaxis.

Functional mapping of the surface of Escherichia coli ribose-binding protein: mutations that affect chemotaxis and transport.

Ribose-binding protein is a bifunctional soluble receptor found in the periplasm of Escherichia coli. Interaction of liganded binding protein with the ribose high affinity transport complex results in the transfer of ribose across the cytoplasmic membrane. Alternatively, interaction of liganded binding protein with a chemotactic signal transducer, Trg, initiates taxis toward ribose. We have generated a functional map of the surface of ribose-binding protein by creating and analyzing directed mutations of exposed residues. Residues in an area on the cleft side of the molecule including both domains have effects on transport. A portion of the area involved in transport is also essential to chemotactic function. On the opposite face of the protein, mutations in residues near the hinge are shown to affect chemotaxis specifically.

Structural and functional analyses of the repressor, RbsR, of the ribose operon of Escherichia coli.

The DNA sequence encoding the rbs repressor protein, RbsR, has been determined. Amino acid sequence analyses of the product of an rbsR-lacZ fusion and of affinity-purified RbsR demonstrate that translation begins at an unusual codon, TTG, and that the initial amino acid is removed during maturation of the protein. DNA-binding assays indicate that RbsR binds to a region of perfect dyad symmetry spanning the rbs operon transcriptional start site and that the affinity for the rbs operator is reduced by addition of ribose, consistent with ribose being the inducer of the operon. RbsR is a member of a family of homologous repressor proteins having very similar DNA-binding sites and binding to similar operator sequences.

Structural homology between rbs repressor and ribose binding protein implies functional similarity.

The deduced amino acid sequence of the rbs repressor, RbsR, of Escherichia coli is homologous over its C-terminal 272 residues to the entire sequence of the periplasmic ribose binding protein. RbsR is also homologous to a family of bacterial repressor proteins including LacI. This implies that the structure of the repressor consists of a two-domain binding protein portion attached to a DNA-binding domain having the four-helix structure of the LacI headpiece. The implications of these relationships to the mechanism of this class of repressors are discussed.

A novel cereal storage protein: molecular genetics of the 19 kDa globulin of rice.

A lambda gt11 cDNA library, constructed from poly(A)+ RNA isolated from immature rice seed endosperm, was screened with affinity-purified antibodies against the rice storage protein called alpha-globulin (previously), or the 19 kDa globulin (our term). A positive clone was isolated and sequenced and shown to encode a 21 kDa precursor for the 19 kDa globulin, based on the identity of portions of the inferred amino acid sequence and the sequence of three cyanogen bromide peptides of the 19 kDa globulin. Analysis of genomic DNA by Southern blotting using the cDNA clone probe revealed one hybridizing band in Eco RI, Hind III, and Bam HI digests. This strongly suggests that the 19 kDa globulin is encoded by a single-copy gene. Because of its single-copy nature and its abundance of Arg and lack of Lys, the 19 kDa rice globulin appears to be a particularly attractive target for genetically engineering increased Lys content in rice seeds.

Cloning and structural characterization of porcine heart aconitase.

A full-length cDNA encoding porcine heart aconitase was derived from lambda gt10 recombinant clones and by amplification of the 5' end of the mRNA. The 2700-base pair (bp) cDNA contains a 29-bp 5' untranslated region, a 2343-bp coding segment, and a 327-bp 3' untranslated region. The porcine heart enzyme is synthesized as a precursor containing a mitochondrial targeting sequence of 27 amino acid residues which is cleaved to yield a mature enzyme of 754 amino acids, Mr = 82,754, having a blocked amino terminus. The NH2-terminal pyroglutamyl residue of the mature enzyme was identified by fast atom bombardment mass spectrometry and sequence analyses of an NH2-terminal peptide. Mature porcine heart aconitase contains 12 cysteine residues. Cysteines 358, 421, and 424 are ligands to the Fe-S cluster in the inactive [3Fe-4S] (Robbins, A. H., and Stout, C. D. (1989) Proteins 5, 289-312) and active [4Fe-4S] (Robbins, A. H., and Stout, C. D. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 3639-3643) forms. An alignment of the derived porcine heart sequence with 8 cysteine-containing tryptic peptides from bovine heart aconitase (Plant, D. W., and Howard, J. B. (1988) J. Biol. Chem. 263, 8184-8189; Plank, D. W., Kennedy, M. C., Beinert, H., and Howard, J. B. (1989) J. Biol. Chem. 264, 20385-20393) shows that 198 of 202 amino acids are conserved and suggests that the two enzymes are virtually identical.

The amino acid sequence of cytochrome c553 from Microcystis aeruginosa.

Cytochrome c553 is an electron donor to P700 in the photosynthetic electron transfer chain of cyanobacteria and eukaryotic algae. We have purified this cytochrome from the cyanobacterium Microcystis aeruginosa and determined its amino acid sequence. When the amino acid sequence of this protein is compared to sequences of cytochromes c553 from other organisms, one sees that the evolution of net charge is more pronounced than the evolution of overall structure, further documenting a pronounced shift in the isoelectric point of this protein during the evolution of cyanobacteria. Cyanobacteria and algae also contain cytochrome c550 (Mr 15,500) which is quite different from cytochrome c553 (Mr 10,500). When the amino acid sequence of cytochrome c553 is compared to that of cytochrome c550, two regions of similar sequence are recognized.

The amino acid sequence of low-potential cytochrome c550 from the cyanobacterium Microcystis aeruginosa.

The low-potential cytochrome c550 has been purified from the cyanobacterium Microcystis aeruginosa and its amino acid sequence has been determined. The protein contains 135 amino acid residues with the Cys-X-X-Cys-His heme binding site at residues 37 to 41. The sequence from residue 28 to 45 shows similarity to cytochrome c553 residues 1 to 18 when the heme binding sites are aligned. Another region of similarity is in the carboxyl-terminal regions of these two proteins. The two aligning regions of cytochrome c553 correspond to helical segments in other related cytochromes. A partial sequence of cytochrome c550 from Aphanizomenon flos-aquae was obtained and showed a 48% identity to the sequence of the M. aeruginosa cytochrome. The single methionine residue in cytochrome c550 of M. aeruginosa occurs at position 119 but there is no methionine in this region in the A. flos-aquae cytochrome, indicating that methionine is not the sixth ligand to the heme iron atom. Histidine 92 is a possible sixth ligand in M. aeruginosa cytochrome c550. The far-uv circular dichroism spectrum indicates that this protein is approximately 17% alpha helix, 42% beta-pleated sheet, and 41% random coil.

Drosophila insulin degrading enzyme and rat skeletal muscle insulin protease cleave insulin at similar sites.

Insulin degradation is an integral part of the cellular action of insulin. Recent evidence suggests that the enzyme insulin protease is involved in the degradation of insulin in mammalian tissues. Drosophila, which has insulin-like hormones and insulin receptor homologues, also expresses an insulin degrading enzyme with properties that are very similar to those of mammalian insulin protease. In the present study, the insulin cleavage products generated by the Drosophila insulin degrading enzyme were identified and compared with the products generated by the mammalian insulin protease. Both purified enzymes were incubated with porcine insulin specifically labeled with 125I on either the A19 or B26 position, and the degradation products were analyzed by HPLC before and after sulfitolysis. Isolation and sequencing of the cleavage products indicated that both enzymes cleave the A chain of intact insulin at identical sites between residues A13 and A14 and A14 and A15. Sequencing of the B chain fragments demonstrated that the Drosophila enzyme cleaves the B chain of insulin at four sites between residues B10 and B11, B14 and B15, B16 and B17, and B25 and B26. These cleavage sites correspond to four of the seven cleavage sites generated by the mammalian insulin protease. These results demonstrate that all the insulin cleavage sites generated by the Drosophila insulin degrading enzyme are shared in common with the mammalian insulin protease. These data support the hypothesis that there is evolutionary conservation of the insulin degrading enzyme and further suggest that this enzyme plays an important role in cellular function.

Identification of phosphorylation sites in peptides using a gas-phase sequencer.

A simple procedure is described for determining the location of phosphorylation sites in phosphopeptides. The method employs measurement of 32P-labeled inorganic phosphate release during Edman degradation cycles using a gas-phase sequencer. The procedure is based on extracting peptides and inorganic phosphate from portions of the sample filter at strategic cycles in the sequence analysis followed by determination of the relative amounts of phosphate and phosphopeptide. One advantage of this technique is the very high recovery of the phosphate associated with the peptide, 80-97% in this study. In the course of this work, it was also found that phosphoserine residues themselves caused reduced efficiency of the Edman degradation as compared with unesterified serine residues. The present procedure has the merit of being simple and easy to apply.

Dynamic properties of membrane proteins: reversible insertion into membrane vesicles of a colicin E1 channel-forming peptide.

The binding of colicin E1 and its COOH-terminal channel-forming peptides to artificial membrane vesicles has an optimum at acidic pH values. The Mr 18,000 thermolytic peptide inserted into membrane vesicles at pH 4.0 has a limited accessibility to exogenous protease. It is converted by trypsin cleavage after Lys-381 and Lys-382 to a lower Mr 14,000 peptide. However, when the pH of a vesicle suspension to which peptide has been bound at pH 4.0 is shifted to 6.0, the accessibility to protease increased greatly. This was shown (i) by the large decrease in the amount of Mr 14,000 or Mr 18,000 peptide after the pH 4----6 shift and treatment with trypsin or Pronase, consistent with (ii) a previously observed decrease in membrane-bound radiolabeled peptide after protease treatment. (iii) When a photoactivable nitrene-generating phospholipid probe was used to label the colicin peptide inserted into the bilayer, the extent of labeling decreased by a factor of 3 when the pH was shifted from 4.0 to 6.5. (iv) Colicin peptide added to vesicles at pH 4.0 can "hop" to other vesicles if the pH and ionic strength of donor vesicles are successively increased. It is proposed that deprotonation of acidic residues in contact with the hydrophobic bilayer or the membrane surface destabilizes the inserted channel and causes it to be extruded from the membrane. The pH-dependent extrusion of the inserted colicin channel provides an example of dynamic properties of an intrinsic membrane protein.

Structure of the coding sequence and primary amino acid sequence of acetyl-coenzyme A carboxylase.

Acetyl-coenzyme A carboxylase (Ac-CoA carboxylase; EC 6.4.1.2) catalyzes the rate-limiting reaction in long-chain fatty acid biosynthesis. To investigate the mechanism of genetic control of expression of Ac-CoA carboxylase and the relationship between its structure and function, cDNA clones for Ac-CoA carboxylase were isolated. The complete coding sequence contains 7035 bases; it encodes a polypeptide chain of 2345 amino acids having a Mr of 265,220. The sequences of several CNBr peptides of Ac-CoA carboxylase were localized within the predicted protein sequence as were those peptides that contain the sites for phosphorylation. The deduced protein contains one putative site for biotinylation in the NH2-terminal half. The "conserved" biotinylation site peptide, Met-Lys-Met, is preceded by valine, whereas alanine is found in a similar position in all other known biotin-containing proteins. The primary sequences of Ac-CoA carboxylase and carbamoyl phosphate synthetase exhibit substantial identity.

Primary structure of soybean lipoxygenase L-2.

The nucleotide sequence of soybean lipoxygenase-2 cDNA has been determined, and the complete amino acid sequence of the enzyme has been deduced. Limited direct amino acid sequence data for lipoxygenase-2 protein support this assignment and exclude mRNA representing lipoxygenase-1 and -3. Lipoxygenase 2 has a molecular weight of 97,036 and contains 865 amino acid residues, in contrast to the isozymes, lipoxygenase-1 and -3, which are known to contain 838 and 859 amino acid residues, respectively. Despite significant differences in behavior between these three isozymes, the amino acid sequences of lipoxygenase-1 and -3 are 81 and 74% identical to lipoxygenase-2, respectively. A region of 40 amino acid residues containing a cluster of six histidines and two tyrosines, which is highly conserved in all three isozymes, is discussed as a possible iron-binding region.

The active form of the cytochrome d terminal oxidase complex of Escherichia coli is a heterodimer containing one copy of each of the two subunits.

The cytochrome d complex is a component of the aerobic respiratory system of Escherichia coli. The enzyme functions as a terminal oxidase, oxidizing ubiquinol-8 within the cytoplasmic membrane and reducing oxygen to water. The enzyme is of particular interest because it is a coupling site in the electron transfer chain. The electron transfer reaction catalyzed by this enzyme is coupled to the translocations of protons across the membrane (H+/e-approximately equal to 1). The oxidase contains two subunits by sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis, with molecular weights of 58,000 and 43,000. In this paper, the question of the quaternary structure is addressed. Quantitative N-terminal analysis of the isolated enzyme and relative mass quantitation following sodium dodecyl sulfate-polyacrylamide gel electrophoresis indicate the subunits are present in equimolar amounts. Sedimentation velocity and sedimentation equilibrium studies were used to characterize the hydrodynamic properties of the purified enzyme solubilized in Triton X-100, under conditions where the enzyme is active. It is concluded that the active enzyme in Triton X-100 is a heterodimer, containing one copy of each subunit. This is likely the structure of the enzyme in the E. coli membrane.

Mutational analysis of the glutamine phosphoribosylpyrophosphate amidotransferase pro-peptide.

Bacillus subtilis glutamine phosphoribosylpyrophosphate amidotransferase is synthesized as a pro-enzyme having an 11-amino acid leader. Maturation requires insertion of a [4Fe-4S] cluster and processing of the pro-peptide to expose an NH2-terminal active site cysteine residue. Point and deletion mutations were constructed in the leader region. These mutations affect processing and enzyme activities. Processing of the leader is dependent upon glutamic acid residues at positions -2 and -1 as well as Cys1. In addition, processing requires a pro-peptide longer than 3 residues. Function of the active site cysteine is dependent on pro-peptide processing. Enzyme purified from a pro-peptide deletion strain has activity and iron content that is comparable to the wild type. These results establish that the pro-peptide is not essential for enzyme maturation, but they leave unanswered the question of pro-peptide function.

Degradation products of insulin generated by hepatocytes and by insulin protease.

The enzymatic mechanisms for insulin breakdown by hepatocytes have not been established, nor have the degradation products been identified. Several lines of evidence have suggested that the enzyme insulin protease is involved in insulin degradation by hepatocytes. To identify the products of insulin generated by insulin protease and to compare them with those produced by hepatocytes, we have incubated insulin specifically iodinated at either the B-16 or the B-26 tyrosines with insulin protease and with isolated hepatocytes, separated the products on high performance liquid chromatography (HPLC), and identified the B-chain cleavages. Insulin-sized products were obtained by Sephadex G-50 filtration. These insulin-sized products were injected on reverse-phase HPLC, and the peaks of radioactivity were identified. The product patterns generated by the enzyme and by hepatocytes were essentially identical with both isomers. The products were also sulfitolized to prepare the S-sulfonate derivatives of the B-chain and B-chain peptides. Again, the patterns on HPLC generated by the enzyme and by hepatocytes with both isomers were identical. Each of the original product peaks was also sulfitolized and injected separately on HPLC to relate B-chain peptides with product peaks. Again, the peptide compositions of the product peaks for both enzyme and hepatocytes were essentially identical. To identify the cleavage sites in the B-chain of insulin produced by insulin protease, the peptides from the degradation of [125I]iodo(B-26)insulin were purified and submitted to automated Edman degradation to identify the cycle in which radioactivity appeared. Seven peptides with cleavages on the amino side of the B26 residue were identified, and the cleavage sites were determined. Cleavages were found between B-9 and B-10 (Ser-His), B-10 and B-11 (His-Leu), B-14 and B-15 (Ala-Leu), B-13 and B-14 (Glu-Ala), B-16 and B-17 (Tyr-Leu), B-24 and B-25 (Phe-Phe), and B-25 and B-26 (Phe-Tyr). Peptides were also isolated from [125I]iodoinsulin incubated with isolated hepatocytes, and the cleavage sites in several of these were determined. These agreed exactly with the cleavage sites identified generated by the enzyme. The major peptides generated by the degradation of [125I]iodo(B-16)insulin were also isolated and sequenced, again showing identical cleavage sites.(ABSTRACT TRUNCATED AT 400 WORDS)

Characterization of osmotin : a thaumatin-like protein associated with osmotic adaptation in plant cells.

Cultured tobacco (Nicotiana tabacum var Wisconsin 38) cells adapted to grow under osmotic stress synthesize and accumulate a 26 kilodalton protein (osmotin) which can constitute as much as 12% of total cellular protein. In cells adapted to NaCl, osmotin occurs in two forms: an aqueous soluble form (osmotin-I) and a detergent soluble form (osmotin II) in the approximate ratio of 2:3. Osmotin-I has been purified to electrophoretic homogeneity, and osmotin-II has been purified to 90% electrophoretic homogeneity. The N-terminal amino acid sequences of osmotins I and II are identical through position 22. Osmotin-II appears to be much more resistant to proteolysis than osmotin-I. However, it cross-reacts with polyclonal antibodies raised in rabbits against osmotin-I. Osmotin strongly resembles the sweet protein thaumatin in its molecular weight, amino acid composition, N-terminal sequence, and the presence of a signal peptide on the precursor protein. Thaumatin does not cross-react with antiosmotin. An osmotin solution could not be detected as sweet at a concentration at least 100 times that of thaumatin which could be detected as sweet. Immunocytochemical detection of osmotin revealed that osmotin is concentrated in dense inclusion bodies within the vacuole. Although antiosmotin did not label organelles, cell walls, or membranes, osmotin appeared sparsely distributed in the cytoplasm.